#pypy on 2019-06-06 — irc logs at freenode.irclog.whitequark.org

2018-11-25 07:18 arigato changed the topic of #pypy to: PyPy, the flexible snake (IRC logs: https://quodlibet.duckdns.org/irc/pypy/latest.log.html#irc-end ) | use cffi for calling C | mac OS and Fedora are not Windows

00:06 jcea has joined #pypy

00:14 adamholmberg has quit [Remote host closed the connection]

00:15 adamholmberg has joined #pypy

00:39 <bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/4647 [default]

00:58 speeder39_ has joined #pypy

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/4648 [py3.6]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/own-linux-x86-32/builds/6475 [py3.6]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/5253 [py3.6]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-macosx-x86-64/builds/4428 [py3.6]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-64/builds/6251 [py3.6]

01:00 <bbot2> Started: http://buildbot.pypy.org/builders/own-linux-x86-64/builds/7501 [py3.6]

01:28 xcm has quit [Read error: Connection reset by peer]

01:29 xcm has joined #pypy

01:36 jcea has quit [Quit: jcea]

01:38 <bbot2> Failure: http://buildbot.pypy.org/builders/own-win-x86-32/builds/2057 [default]

02:05 <bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/5253 [py3.6]

02:27 <bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-64/builds/6251 [py3.6]

02:30 dddddd has quit [Remote host closed the connection]

02:48 oberstet_ has quit [Remote host closed the connection]

03:18 speeder39_ has quit [Quit: Connection closed for inactivity]

03:38 <bbot2> Failure: http://buildbot.pypy.org/builders/own-linux-x86-32/builds/6475 [py3.6]

03:40 <bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-macosx-x86-64/builds/4428 [py3.6]

03:40 <mattip> another project now has pypy running in CI - https://github.com/scikit-build/scikit-build/pull/418

03:41 <mattip> scikit-build wraps cmake and is needed to build opencv

03:41 <mattip> python-opencv

03:56 _whitelogger has joined #pypy

04:02 <bbot2> Failure: http://buildbot.pypy.org/builders/own-linux-x86-64/builds/7501 [py3.6]

05:00 <bbot2> Started: http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-64/builds/2617

05:06 <cfbolz> mattip: nice!

06:28 Masklinn has joined #pypy

06:32 tsaka__ has quit [Ping timeout: 272 seconds]

06:47 xcm has quit [Remote host closed the connection]

06:49 xcm has joined #pypy

06:53 <bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-win-x86-32/builds/4648 [py3.6]

07:07 oberstet has joined #pypy

07:08 inhahe has quit []

07:10 inhahe has joined #pypy

07:15 <bbot2> Success: http://buildbot.pypy.org/builders/jit-benchmark-linux-x86-64/builds/2617

07:35 mattip has quit [Ping timeout: 244 seconds]

07:38 themsay has quit [Read error: Connection reset by peer]

07:39 themsay has joined #pypy

07:57 tsaka__ has joined #pypy

08:02 mattip has joined #pypy

08:04 <bbot2> Started: http://buildbot.pypy.org/builders/jitbackendonly-own-linux-armhf-v7/builds/1551 [default]

08:59 <bbot2> Retry: http://buildbot.pypy.org/builders/jitbackendonly-own-linux-armhf-v7/builds/1551 [default]

08:59 <bbot2> Started: http://buildbot.pypy.org/builders/jitbackendonly-own-linux-armhf-v7/builds/1552 [default]

09:01 mattip has quit [Ping timeout: 258 seconds]

09:05 themsay has quit [Ping timeout: 246 seconds]

09:18 jacob22 has quit [Read error: Connection reset by peer]

09:21 antocuni has joined #pypy

09:21 jacob22 has joined #pypy

09:21 mattip has joined #pypy

09:24 rubdos_ is now known as rubdos

09:35 themsay has joined #pypy

09:59 themsay has quit [Read error: Connection reset by peer]

10:00 themsay has joined #pypy

10:02 mvantellingen has quit [Ping timeout: 252 seconds]

10:10 pierreg has joined #pypy

10:11 mvantellingen has joined #pypy

10:16 <bbot2> Retry: http://buildbot.pypy.org/builders/jitbackendonly-own-linux-armhf-v7/builds/1552 [default]

10:22 mvantellingen has quit [Ping timeout: 252 seconds]

10:22 pierreg has quit [Quit: Page closed]

10:22 pierreg has joined #pypy

10:45 xcm has quit [Remote host closed the connection]

10:47 xcm has joined #pypy

10:47 dddddd has joined #pypy

11:02 jamadden has joined #pypy

11:09 antocuni has quit [Ping timeout: 272 seconds]

11:22 mvantellingen has joined #pypy

11:23 jcea has joined #pypy

11:29 <bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-linux-s390x/builds/1085 [default]

11:46 kajetanj has joined #pypy

11:46 kajetanj has left #pypy [#pypy]

11:46 kajetanj has joined #pypy

11:50 kajetanj has quit [Client Quit]

11:58 adamholmberg has quit [Remote host closed the connection]

11:59 adamholmberg has joined #pypy

12:03 adamholmberg has quit [Ping timeout: 248 seconds]

12:18 lritter has joined #pypy

12:29 antocuni has joined #pypy

13:00 Masklinn has quit [Ping timeout: 246 seconds]

13:00 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-lib-python-linux-x86-64/builds/57 [cfbolz: force build, json-decoder-maps]

13:00 <bbot2> Started: http://buildbot.pypy.org/builders/own-linux-x86-32/builds/6476 [cfbolz: force build, json-decoder-maps]

13:00 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/5254 [cfbolz: force build, json-decoder-maps]

13:00 <bbot2> Started: http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-64/builds/6252 [cfbolz: force build, json-decoder-maps]

13:00 <bbot2> Started: http://buildbot.pypy.org/builders/own-linux-x86-64/builds/7502 [cfbolz: force build, json-decoder-maps]

13:02 Masklinn has joined #pypy

13:02 jcea has quit [Remote host closed the connection]

13:02 <bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-32/builds/5254 [cfbolz: force build, json-decoder-maps]

13:06 adamholmberg has joined #pypy

13:11 jcea has joined #pypy

13:17 pierreg has quit [Quit: Page closed]

13:33 <kenaan> mattip py3.6 b0cb066e7e1f /pypy/: trivial fixes

13:33 Rhy0lite has joined #pypy

13:45 <bbot2> Failure: http://buildbot.pypy.org/builders/own-linux-x86-32/builds/6476 [cfbolz: force build, json-decoder-maps]

13:51 <bbot2> Success: http://buildbot.pypy.org/builders/pypy-c-lib-python-linux-x86-64/builds/57 [cfbolz: force build, json-decoder-maps]

14:05 <bbot2> Failure: http://buildbot.pypy.org/builders/own-linux-x86-64/builds/7502 [cfbolz: force build, json-decoder-maps]

14:08 <arigato> issue 3014: "how could it ever have worked"

14:08 <arigato> I think it's caused by the following optimization pass:

14:10 <arigato> p95 = new_array(4, descr=<ArrayU 1>)

14:10 <arigato> # copy 4 bytes into this new array:

14:10 <arigato> # read the initial two bytes as a single operation:

14:10 <arigato> i127 = gc_load_indexed_i(p95, 0, 1, 16, 2)

14:10 <arigato> call_n(ConstClass(ll_arraycopy__arrayPtr_arrayPtr_Signed_Signed_Signed), p96, p95, 0, 0, 4, descr=<Callv 0 rriii EF=2 OS=1>)

14:10 <arigato> gets optimized as:

14:10 <arigato> p182 = new_array(4, descr=<ArrayU 1>)

14:10 <arigato> i235 = gc_load_indexed_i(p182, 0, 1, 16, 2)

14:10 <arigato> setarrayitem_gc(p182, 0, i183, descr=<ArrayU 1>)

14:10 <arigato> setarrayitem_gc(p182, 1, i186, descr=<ArrayU 1>)

14:10 <arigato> setarrayitem_gc(p182, 2, i188, descr=<ArrayU 1>)

14:10 <arigato> setarrayitem_gc(p182, 3, i190, descr=<ArrayU 1>)

14:10 <arigato> the load has been moved before the four setarrayitem_gc's

14:11 <arigato> or more precisely the setarrayitem_gc have been delayed and occur later in the trace

14:13 Graypup_ has quit [Quit: ZNC 1.6.1 - http://znc.in]

14:14 <mattip> strange

14:14 <mattip> since what you have here definitely looks wrong

14:15 Graypup_ has joined #pypy

14:17 <antocuni> weird

14:17 <antocuni> I am using struct.unpack extensively in capnpy and in gambit we have lots of processing reading them 7/24, never found any error

14:17 <arigato> it's clearly what occurs, I could write an optimizeopt test

14:19 <arigato> note that in pypy, the problem only seems to occur by struct.unpack(.., bytearray(..)[:slice])

14:19 <antocuni> ah I see, it's the slice which produces the ll_arraycopy

14:19 <arigato> yes, but for some reason the optimizeopt test doesn't require ll_arraycopy

14:19 <antocuni> ok, this is a pattern I never use, so it explains why I never encountered the bug

14:20 <arigato> maybe it's a pattern where you build a bytearray somehow and then pass it to struct.unpack()

14:24 <arigato> ah, if you manipulate a string (and not a bytearray) then you get lots of copystrcontent, which are not reordered

14:31 <bbot2> Failure: http://buildbot.pypy.org/builders/pypy-c-jit-linux-x86-64/builds/6252 [cfbolz: force build, json-decoder-maps]

14:40 Masklinn has quit [Ping timeout: 246 seconds]

14:41 irclogs_io_bot has quit [Remote host closed the connection]

14:48 <kenaan> arigo default 4570c3e9b030 /rpython/jit/metainterp/optimizeopt/: Issue #3014 The gc_load family of operations must force the lazy setfields and setarrayitems to occur first

14:50 <kenaan> arigo default 7419bb5bbb41 /rpython/jit/metainterp/optimizeopt/heap.py: meh, fix for 4570c3e9b030

14:52 irclogs_io_bot has joined #pypy

14:54 <kenaan> arigo default 631c5b49bd25 /rpython/jit/metainterp/optimizeopt/test/test_optimizeopt.py: Improve the test, still passing

14:57 Masklinn has joined #pypy

15:14 Masklinn has quit []

15:41 <cfbolz> arigato: ouch

15:41 <cfbolz> That's really terrible

17:00 moei has joined #pypy

17:15 antocuni has quit [Ping timeout: 248 seconds]

17:28 jamadden has quit [Quit: Leaving.]

17:29 <arigato> yes, bad us

17:37 jacob22 has quit [Read error: Connection reset by peer]